parser: add nowait keyword, add for update no wait grammar, and related error…

[cfzjywxk]

… code/msg

What problem does this PR solve?

support select for update nowait

# Session 1:

mysql> CREATE TABLE t (i INT, PRIMARY KEY (i)) ENGINE = InnoDB;

mysql> INSERT INTO t (i) VALUES(1),(2),(3);

mysql> START TRANSACTION;

mysql> SELECT * FROM t WHERE i = 2 FOR UPDATE;
+---+
| i |
+---+
| 2 |
+---+

# Session 2:

mysql> START TRANSACTION;

mysql> SELECT * FROM t WHERE i = 2 FOR UPDATE NOWAIT;
ERROR 3572 (HY000): Statement aborted because lock(s) could not be acquired immediately and NOW

What is changed and how it works?

Check List

Tests

• Unit test

Code changes

• Has exported variable/fields change

Side effects

Related changes

• Need to update the documentation

[codecov]

Codecov Report

Merging #582 into master will increase coverage by <.01%.
The diff coverage is 75%.

@@            Coverage Diff             @@
##           master     #582      +/-   ##
==========================================
+ Coverage   71.82%   71.83%   +<.01%     
==========================================
  Files          32       32              
  Lines        7834     7836       +2     
==========================================
+ Hits         5627     5629       +2     
  Misses       1676     1676              
  Partials      531      531

Impacted Files	Coverage Δ
parser.go	70.37% <ø> (ø)	⬆️
misc.go	96.49% <ø> (ø)	⬆️
ast/util.go	75% <0%> (ø)	⬆️
ast/dml.go	73.92% <100%> (+0.03%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 448f044...59c18d8. Read the comment docs.

[kennytm]

From the description in https://dev.mysql.com/doc/refman/8.0/en/select.html#idm140006656103232, it's more appropriate to make NOWAIT its own modifier rather than grouping FOR UPDATE NOWAIT as a whole thing.

    [FOR {UPDATE | SHARE} [OF tbl_name [, tbl_name] ...] [NOWAIT | SKIP LOCKED] 
      | LOCK IN SHARE MODE]]

[coocood]

NOWAIT is an unreserved keyword, we need to add it to UnReservedKeyword: section in the parser.y.

[cfzjywxk]

From the description in https://dev.mysql.com/doc/refman/8.0/en/select.html#idm140006656103232, it's more appropriate to make NOWAIT its own modifier rather than grouping FOR UPDATE NOWAIT as a whole thing.

    [FOR {UPDATE | SHARE} [OF tbl_name [, tbl_name] ...] [NOWAIT | SKIP LOCKED] 
      | LOCK IN SHARE MODE]]

we're currently groupint "FOR UPDATE" as a whole, and "SelectLockType" is defined as single int type, maybe we need to some refactorations on this for 8.0 compatibility ? There will be some other related work, do this refactoration in the future?
@coocood @kennytm

[coocood]

LGTM

[cfzjywxk]

@jackysp @tiancaiamao PTAL

[kennytm]

we're currently groupint "FOR UPDATE" as a whole, and "SelectLockType" is defined as single int type, maybe we need to some refactorations on this for 8.0 compatibility ? There will be some other related work, do this refactoration in the future?

@cfzjywxk IMO it's better to turn SelectLockType into a struct now, or add a new member to SelectStmt indicating whether it is NOWAIT (and SKIP LOCKED in the future).

To support SelectLockForUpdateNoWait you'll need to modify TiDB once now, and later when you refactor for 8.0 compatibility you'll need to modify the code again. So I think it's better to choose a solution which we know is going to last.

[coocood]

@kennytm
We can modify the code again later when we need to.
It's not certain if we need to support those features.

[jackysp]

LGTM